VGA and USB status report

Matan Ziv-Av matan@svgalib.org
Tue Jun 29 15:01:53 CEST 2004


On Tue, 29 Jun 2004, Oleg Gusev wrote:


I found the problem that caused the panic - there is a workaround for 
strongarm bug. I simply removed it.

hid - works (only mouse tested).

usb-storage - works (only CF reader tested).

usbnet (USB-USB cable) crashes on ifconfig. (actually, sometimes it does 
not crash, and arps go both ways fine, but no ping requests or replies).

rtl8150 (USB-ethernet bridge) crashes on ifconfig.


> > usb-ohci.c: USB OHCI at membase 0x0, IRQ 101
> > hc_alloc_ohci: dev=457 mem_base=00000000 dev_is_h5400=0
> 
> mem_base=0 is bad. It is followed by ohci->regs=mem_base.
> In SA1111 it is USB_OHCI_OPBASE.

The ohci code assumes that ohci registers are packed in memory, bu 
sa1101 has them 1KB apart, so I redefined writel (see start of 
usb-ohci). This means that mem_base is meaningless.

> > hcca=c1af0000  hcca_dma=c0046000
> 
> Have you reserved the lowest 2MB for DMA like SA1111 does it ?

No. should I? Is there a known bug with sa1101 DMA?


> Have you enabled DMA in TUCR ? The resource handling
> should be done properly sometime, but as a quick test
> it will go.

Yes.

> > usb.c: USB device 2 (vend/prod 0x67b/0x0) is not claimed by any active
> > driver.
> > usb0: register usbnet 001/002, Prolific PL-2301/PL-2302

> PS. You are very brave person to deal with the 2.4 SA USB code :)

Actually, as you can see, there is very little work to do.


-- 
Matan Ziv-Av.                         matan@svgalib.org



More information about the Jornada820 mailing list